<![CDATA[Soil moisture plays a crucial role in plant growth as it affects nutrient transport, photosynthesis, transpiration, and cell turgor. Soil moisture is commonly measured using direct gravimetric methods or more practical electrical conductivity-based sensors. However, these sensors have drawbacks, as measurement results can be affected by salinity, temperature, and soil type, requiring calibration at each site. To address these limitations, this study employed an Artificial Neural Network (ANN) model integrated into a microcontroller to automate the measurement process, adapt it for specific tasks, and eliminate manual calibration. The objectives of this study were to design a soil moisture sensor without manual calibration, develop an ANN-based moisture estimation model, and test its accuracy and effectiveness in supporting precision agriculture. The results showed that the soil moisture meter portable performed according to the design criteria. The ANN model produced an RMSE of 1.0366 and an R² of 0.995 in the training phase, and an RMSE of 1.0366 and an R² of 0.9923 in the testing phase. The best activation function is logsig-logsig-logsig, with actual validation showing an R² of 0.8607 and an RMSE of 7.617. This system has the potential to improve the accuracy of soil moisture measurements and irrigation efficiency in supporting sustainable precision agriculture.]]>
Copyrights © 2026
